TRADITIONAI. SYSTEM OF GOAT MANAGEMENT.
III. REPRODUCTIVE PERFORMANCE OF SRD (NON-DESCRIPT) DOES AND KID MORTALITY1 FRANCISCO DE A. MELO LIMA 2 , AURINO A. SIMPLÍCIO3,
ÉL$IO ANTONIO P. FIGUEIREDO 4 e FEDERICO A. PONCE DE LEÓN 5
ABSTRACT - lhe reproductive performance o! 50 adult SRD does maintained in native caatinga pasture at a stocking rate o! 1 to 1.5 ha/head/year was evaluated. Two bucks were used in natural matings from September 1978 until December 1980. A total o! 93 kiddirigs and 28 abortions oc-curred. Turco pregnancies occurred before initiation of the experiment. From fite 90 parturitions, 57.7%, 35.6%, 1.1% and 5.6% bore singles, twins, tripiets and undetermined, respectively. 'Ite tiock showed an overali proiificacy o! 1.38. From the initíal fiock, 15(30%). 28(56%), 5(10%) and 2(4%) does presented one, two, three and zero kiddings, respectively during the period o! study. Neither gestation iength nor kidding interval were influenced by type o! birth. Length of open period was influenced by season of the year when data were ciassified by season and kidding status at the open period. Combination of kidding type at previous and foliowing kiddings showed a significant influence (1' < 0.05) on the iength of the open period. Animais which kidded first singies and then singles or twins showed a shorter open period than animais which !irstiy kidded twins and then aborted at the following pregnant (109 vs 188 days). 'lhe peak o! Md mortality occurred between 7 and 112 days of age. lCid morta!ity appears not to be influenced by sex or type o! birth. Overali kid mortaiity was 76%. The age at dead o! Idds was influenced by season and sex as weli as by their interaction. Males bern during the rainy season survived longer than any other ciass studied.
Index terms: native breed,reproduction,mortallty,
SISTEMA TRADICIONAL DE MANEJO DE CAPRINOS.
III. DESEMPENHO REPRODUTIVO DE CAPRINOSSRD E MORTALIDADE DE CABRITOS RESUMO . Avaliou-se o desempenho de 50 fêmeas de cabras SRD adultas mantidas em um piquete de pastagem nativa com uma Iotaç5o de 1 a 1,5 hafcab/ano. Neste rebanho, dois reprodutores foram usados em monta natural de setembro de 1978 e dezembro de 1980. Foram observados 93 panos e 28 abortos. Ocorreram três prenhezes antes do Inicio do experimento. Dos 90 panos restantes, 57,1%, 35,6%, 1,1% e 5,6% forem simples, duplos, triplos e nio determinados, respectivamente. O rebanho mostrou uma prolificidade de 1,38. Para o rebanho inicial 15(30%), 28(56%). 5(10%) e 2(4%) das
cabras pariram uma, duas, três e nenhuma vez, respectivamente, durante o período estudado. O
perío-do de gesteçio e o Intervalo entre os partos nio forem influenciados pelo tipo de nascimento. O pe-ríodo vazio foi Influenciado pela estaçio do ano, quando os dados foram classificados por estaçio e tipo de parto. A combinaçio do tipo de parto, na pariçio inicial e na pariçffo seguinte, teve influên-cia significativa no período vazio. Os animais que pariram primeiramente simples e depois simples ou duplos apresentaram um período vazio menor (109 vs 188 dias) do que os animais que pariram pri-meiramente múltiplos e depois abortaram na prenhez seguinte. O pique de mortalidade (32%) ocorreu entre 7 e 112 dies de Idade. A monalidade de cabritos parece nio ter sido Influenciada pelo sexo e tipo de nascimento. A mortalidade total das crias foi de 16%. A idade ô mone desses animais foi in-fluencieda pela estaço, sexo e interaçio de ambos. Os machos nascidos durante a estaçio chuvosa so-breviveram mais tempo do que os nascidos nas outras classes estudadas.
Termos para Indexaçio: raça nativa, reproduç!o, mortalidade de crias, produçio de crias
INTRODUCTION 1
Accepted for publication on November 18fh, 1988. 2
Eng. - Agr., M.Sc., UniversidadeFederal do Ceará - Dep. Zootecnia. Caxnpus do Pici, CEP 60000 Forta-leza, CE.
Méd. - Vet., Ph.D., EMBRAPA/Centro Nacional de Pesquisa de Caprinos (CNPC), Caixa Postal PIO, CEP 62100 Sobra!, CE.
Zootecnista, Ph.D. EMBRAYA/CNPC.
Eng. -Zootecnista, Ph.D., Consu!tant 11C4'EMBRAPA.
One of the major goals of the Braziian National Goat Research Center (CNPC) - EMBRAPA is to develop simple and interrnediate technologies which could be of yalue to small goat producers ia increasing goat fiock productivity.
The net return of an animal is strongly related to the reproductive capacity of the herd (Glimp 1971) since the performance ia that arca, closely Pesq. agropec. bras., Brasúia, 23(12):1449-1458, dez. 1988.
1450 F. DE A.M. LIMAet ai.
associated with nutritional factors, determine : animal base. Improvenzents upon reproductive efficiency, such as identifying and culling infertile does and increase in the nurnber of kids born per kidding within a period of time could be attained with simple husbandry measures. The initiai step lii this direction was taken by evaivating the pre. and post - weaning growth of native goats under traditional system of management (Lima et ai. 1983, 1985). The research described lii this paper intends to quantify the reproductive performance of the fiock and the mortality of kids under thc traditional system.
MATERIALS AND METHODS
Location
This worlc was carried out at the Brazilian National Goat Research Center (CNPC) - EMBRAPA, from Sep-tember 1978 until December 1980. C!imatological data of lhe area have been reported e!sewere (Figueiredo & Pant 1982, Lima etal. 1983).
Animais and management
From the popu!ation of SRD (nondescript) does of lhe CNPC, fifty adult does were chosen at random and maintained within a fenced native. "caatinga" pasture at the stocking rate of 1 to 1.5 ha/head/year. Two bucks were used in natural matings throughout the year. Identi-fication of matings was possib!e by marking lhe bucks on the chest with a mixture aí grease and powder co!or ink, changing co!or every fifteen days. Overali managemenl ofthe herd was described e!sewhere (Lima eI ai. 1983).
Data and statistical methods
From Seplember 1978 to December 1980, a total of 127 kids were bom out of 93 recorded partumitions. Data were oblained on several paramelers as ío!lows: Uve weight of does: Data were recorded every 28day interval. Doe body weights at lhe last two months of lhe gestation period were not considered for calcu!ation of lhe average rnonthly weights.
Geslation Iengtli . At the beginning of the expemimenl, the mating dates were not recorded on every animal. Thus, the mean gestationlength was caicu!ated from 46 observations of which, 27, 18 and 1 gestalion resulted in sing!es, lwins and undetermided lype of births, respecli-ve!y. An analysis of variance (Stee! & Torne 1960) was carried out lo determine the efíect of type of birth on gestation length.
Kidding interval . fle time lapsed between kiddings was calculated from 33 obsenvations. Data were c!assified according to type of birth to tesl the effecl aí kidding
status on kidding interval.
Operi perlod . Def!ned as lhe lime lapsed between the date oí previous kidding and the date aí mating !eading to either a successfull lddding or lo an abortion. Depend-ing on the available number of observalions data were classified accordind to lhe foflowing crileria:
1. By season (rainy and dry), and kidding stalus (single and lwins) of previous kidding, to lest lheir effects on lhe !ength af the open period.
2 >' season of previous kidding and by lhe combina-tions of previous and following kiddings stalus (single to
single or twin, SST; Single lo abortion, SA ;Twin lo single ar twin, SST; and twin to abortion, TA) lo tesl its effecls on the open period interval. Analysis of variance was canied out by lhe method of weighted squares of means (Stell & Torne 1960).
Kidi mortalily Duning the period aí study 97 !dds died. Available data were c!assifiçd according lo month of birth, monlh oí death, sex, lype of birth and age at dealh. Out of 105 kids bonn duning the year 1979, 75 died. Comp!ete data were available on 73 animais. lhe general linear model procedure of !easl squares was used for ana-lysis (Barr eI al. 1979) lo lest lhe fixed effects ofseason, sex and type of birth on death age of kids.
RESULTS
Does'weight
A graph of the mean does'syeigths every 28-days. is presented lia Fig. 1. li can be appreciated that variation of mean weights between mnonths could be related to rainfail pattern and, in so doing, to fodder availabiity. The mean weight peak over 32.0 kg for the years 1979 and 1980 was reached during the month of August, whiie the lowest live weight mean, around 20.0 kg, was commonly reached duning the month ofJanuary. Parturitions and abortions
From September 1978 until December 1980 a total of 93 parturitions and 28 abortions were mecorded, theim monthiy distnibution are presented in Fig. 2. Of the total number of partunition rb-corded, three were from pregnancies which started before the initiation of the experixnent, the remai-ning 90 partunitions produced 124 kids, distni- buted in 57.7%, 35.6%, 1.1% and 5.6% singies, twins, tniplets and undetermined type of births, respectively, showing an overail prolificacy of 1.38. From the initial fiock 15(30%), 28(56%), 5(10%) and 2(4%) does had 1, 2, 3 and zero kid- dings, respectiveiy, duning the period of study. Pesq. agropec. bras., Brasilia, 23(12):1449-1458, dez. 1988.
3, 30 20 300 3 200 3 100
TRADITIONAL SYSTEM OF GOAT MANAGEMENT 1451
MONTHS VEARS
FIG. 1. Woight of does at 28 day jntervals and rainfail pattern.
Out of 13 yearling does w}oich were produced as replacernents, only 4(31%) kidded within the pe-riod of study. Fifty three kiddings and 19 abor-dons occurred during the dry season while 40 kid-dings and 9 abartibns were registered during the rainy season.
Gestation Iength and kidding interval
Type of birth did not affect (P > 0,05) the length of the gestation period (Table 1). Mean gestation lengths were 146.1 and 145.1 days for single and twin gestations, respectively (Table 2). Kidding status, similarly, did not aífect (P > 0,05) kidding interval.
Open period
Analysis of variance and least-square means are presented, respectively in Tables 3 and 4.
When data were classified according to season and kidding type at previous kidding, that is, at the beginning of the open period, the season was found to significantly (P < 0,05) influence the open period inter-vai, this being shorter (124.8 days) when the open period began in the rainy season as opposed to the dry season (170.4 days). Kidding type of the does or the season x kidding status interaction did not influence the open period (1'> 0,05).
Under the criteria of classifying data according to season at previous kidding and combination of kidding type at previous and foliowing kiddings, it was found that combination of kidding type significantly influenced (P < 0,05) the iength of the open period, while season or doe interaction of botli main effects were not important. Does wich Pesq. agropec. bras., Brasilia, 23(12):1449-1458, dez. 1988.
a o
1452 P flPÂKA rIpJA.j
5 O !l O J E M A M J A $ O N O J E M A M J Ju A 5 O N D
MONTPIS VEARS
FIO. 2. Distributjon of parturitions and abartions lii relation to rainfali pattern.
kidded either singles ar twins at the beginning of che opeuperiod and then aborted at the foliowing kidding showed longer open periods than animais which successfully kidded in two consecutive kiddings, although the group af animais kidding twins and then singles ar twins (TST), was not significantly different from the formers. A signifi-cane difference (1' c 0,05) was found between the group aí animais which kidded singies and then singles ar twins (SST) and the group whichíirstiy kidded twins and then aborted in the foliowing kidding (TA).
Kids' mortality
The incidence aí kid moreality according to age is presented in Tabie S. The highest mortality rate (32.3%) occurred for the period between 7 and 112 days af age. Lower morrality rates oc-curred during the Lirse week (18.8%) and pose--weaning (20.5%) períods. However, whel) morta-lity rates were expressed per unit of time (% mortali-ty/day), it was observed that highest incidence of moreality occurred during the íirst three days Pesq. agrapec. bras., Brasília, 23(12):1449-145$, dez. 1988.
of life. The distribution of kid mortaiity with res-pect to sex and type aí birth is presented in Ta-bie 6. li shows no clear cut advantage aí males over females and of singies over twins. Overail kid mortality was racher high, reaching 76.4% morta-hty for the fiock during the period aí study.
The least-squares analysis af age at death af kids (Tabie 7) showed significant (P < 0,05) season and sex effects as weil as season x sex inte-raction. Kids born in che rainy seasan and male kids survived ianger than kids born in the dry season and female kids, respectively. Alsa, male kids born during che rainy season survived ianger chan either females born during the rainy season ar males and females bamn during the dry season. Fig. 3 shaws the distributian aí kid and adule mar-talities as related to the rainfall pattemn thraughout the periad of study. lt can be abserved that the highest mortality occurred during the dry rnonths (August-January) and that mortality is drasticaily reduced as soou as the amaunt aí mainfali increa-ses. Fig. 4 illustrates the reiative advantage af kids born in the rainy seasan as appased to the anes
TRADITIONAL SYSTEM OF GOAT MANAGEMENT 1453 o o -J a o o o a = z 35 30 20 'o NTHS YERS
FIG. 3. Kídsond adult does mortality distribution in relation to rainfail patern.
born in the dry season. A lower percentage of kids born during January and February 1979 (rainy season) died during the course of the experiment, than those born in any other month, when morta-lity rates by month ranked from 90% to 100%.
DI SCUSSI ON
Under the ciimatic conditions of the geographi-cai region and under the constraints imposed to this experiment in trying to imitate as much as possibie the traditional system of management, the season effect has been the most important and significant one influencing ali production para-meters analysed (Lima etai. 1983, 1985).
Doe body weigth was highly correlated to raio pattemn in as much as the onset of the rainy season dictated the triggering of an increase of edible ve-getal material and this in turn triggered a weigth gaining phase. As the edibie material decreased
in quantity and quaiity, asa result of consumption and lack of rains, a weigth losing phase was im-posed ou the animais. in this way, the highest mean weight was reached around August, three months after the cessation of raios. The Iowest mean weight occurred around January, at the onset of the rainy season. The diffeferece between the highest and the iowest mean weights was about 12 kg. Between August and Janaury, the animais showed a weight loss of about 80 g/day, whiie between February and August, the animais were regaining weight at a rate of 57 glday. The weight osciliation should cause some íertility problems for the fiock increasing number of abortion, service periods and corresponding kidding intervais. The kidding pattemn throughout the period of analysis is not regarded as a natural one, since bucks were introduced in a fiock of dry animais ar the beginning of the experiment and this may have stimulated does to s° into estrus (Shelton 1960). Thus, 39 pregnancies started in September Pesq. agropec. bras., Brasília, 23(12):1449-1458,dez. 1988.
IA a lá. o tu 2
1454 E. DE A.M. LIMA et ai.
J F M A M J ju A 5 O N D J F M A M J Ju A 5 O N D
MONTHS YEARS
Figuro 4. Number o? Kds bom within rnonth and number of those kids which die during lhe
course of lhe experimen?.
FIG. 4. Number of Idds bern within month and number of those kids which die during the course of the experiment.
TABLE 1. Analyses of variance of gestation Iength (days) and kidding interval (days) of SRI) goats.
Source of
d f Gestation Kidding
variation length interval
Type of birth 1 10.79 ns 7033.97 ns
Errar § 9.73 7626.55
(43) (28)
§ Degrees ol treedorn within parenthesis.
TABLE 2. Means of gestation length and kidding interval
SRI) goats.
Main
Classes Gestation Kidding
etfect Iength (days) interval (days)
Type of Singles 146.11 (0.121)3 274.18 (22.72) a
of birth Twins 145.11(0.419) a 305.08(21.65)3
Means with the sarne superscrlpts are not slgníficantly differént (P 0,05).
TRADITIONAL SYSTEM OF GOAT MANAGEMENT 1455
TABLE 3. Analyses of variance of sente period (days) according te clifferent criteria of classification of data 1.
Soason and kidding Season at the status at the beginning of the beginning of the service period and Source of service period combination of
variation kidding status at
previous and kiddings d.f. MS d.f. MS Season (S) 1 20454.84 1 10738.38 ns Kidding 1 9235.20 ns 3 13072.19 Status (K) SxK 1 561.47ns 3 1071.81ns Error 41 4205.47 37 3967.76
Analyses of variance by the method of weighted means.
TABLE 4. Unweighted means of sente period Iength (days) as affected by season and kidding status.
Season and Season at the kidding status beginning of the ser- at the begin- vice and combination Main ning of the of kidding status at eflects Classes service previous and follow-
period Ing kiddings
Season Rainy 124.8a 134.1 a
Dry 170.4b 170.0a lCidding Single 132.3 o - Status twin 162.9 e - SST - 180.7b SA - 174.7bc TST - 137.0bc TA - 187.8c
Note: Comparisons were mado within column and within main ef!ects;
dif!erent superscripts show a signi!Icant differenco (P 0,05);
SST: single to single or twins. SA: singles to abortion. TST: twin to single or twins. TA: twin to abortion.
and October 1978 of which four resuited in abor. dons. The first kidding peak coincided with the
onset of the rainy season, and the second peak
occurred during the dry season, between August and October 1979.
TABLE S. Kid mortality according te age.
Mortality percent
Number Percentage age based Day Age of kids of kids of total on total 1%)
mortality number of kids bom 0-24(h). 9 9.3 7.1 9.3 25-48 (h) 7 7.2 5.5 7.2 49-72 (h) 4 4.1 3.1 4.1 4. 7(d)b 4 4.1 3.1 1.4 7 - 112(d) 41 42.3 32.3 0.4 112(d) 26 26.8 20.5 - N.l. 6 6.2 4.7 - Note: a -hours. b days. Nt. • Not identified.
TÀELE 6. 1CM mostatity according to sex and type of birth.
Sex Type o! Total Total Percent of birti, bom deaths mortality
Male Single 28 22 78.6 Twin 30 19 63.3 Female Single 27 20 74.1 Twin 34 27 79.4 - Triplets 3 3 100.0 Ni. - 5 6 - Total - 127 97 764 W. • Not identified
li is expected for a doe to have at ieast one kidding per year uniess reproductive probiems
exist. From the 50 does used in the experiment,
oniy two did not kid during the period of the study and could be regarded as sterile ones. The remaining 48 does shouid have given at ieast 96 kiddings Nevertheiess, oniy 33(69%) animais
kidded twice and 5(15%) of these, three times,
within the period anaiyzed. Fifteen (31%) does
out of the 48 fertile had oniy one kidding in two
years, showing a clear reproductive inefficiency.
1456 F. DE A.M. LIMA etal.
TABLE 7. Lest-squaxes analysis of variance and Iest-squares means of age at death (days) of SRD kids. Source of
variatuon d.f. - Mean squares
Least-square means
Classes X (s.e)
Season (E) 1 46683.59 Rainy 139.6 (25.8)8
Dry 80.0(14.1)b
Sex (5) 1 54608.78' Male 141.0 (18.5) c
Female 78.7 (22.1) d
Type of birth 1 3682.66 ns Single 102.1 (21,1) e
Multiple 117.5(18.6)0 Rainy -males 199.8 (28.6) Ex 5 1 48656.73' Rainy - females 79.5(41.4)9 Dry 'males 82.2 (22.6) 9 Dry -females 77.8(17.7)9 ExT 1 31.99ns - - Regression on 1 35892.80 a - - birth weigth Error 80 11612.80 - - * (P 0,05). a (P - 0,0826). ns = non-significant.
Means witli sarne superscript withln main classification are not significantly different (P 0,05).
Feed was scarse during the dry season and ani-mais may have had to survive on a iow energy pian, which may have contributed for the high abortion rate. The effect of nutrition plan on re-productive efficiency had been demonstrated on does fed isonitrogenous diets containing low-ener-gy leveis (Sachdeva et al. 1973).
Gestation lenght and kidding interval were not influeced by type of birth nor kidding type, res-pectiveiy. This is in agreemént with the findings of Mishra et a! (1979) for the Sirohi and Beetal, Gupta et ai. (1964) for the Biack-Bengal, Kirkpa-trick & Akindele (1974) for the West African Dwarf breeds.
Gestation iength is commonly influenced by breed, birth weight, nutritional plant (Jardim et ai. 1965, Forbes 1967, Sachdeva et al. 1973, Gonzaiez-Stagnaro 1977) and probably by type of birth, although there are some conflicting cvi' dences for the !atter effect. For sheep it lias been reported that muitipie births are associated with shorter gestations (Forbes 1967), whule Giimp (1971) found that the gestation lenght of SuffoUc,
Targhee, Corriedale, Navajo and Coarse wool breeds was not infiuenced by type of birth.
Factors affecting open period in goats have not been thoroughiy investigated. Although with small number of observations, this study was an attempt to do so. When data were c!assified according to season of kidding and kidding type at the begin' ning of the open period, season was found to signi-ficantly influence the open period. Does kidding during the rainy season showed shorter open periods than does kidding during the dry season, the season effect being correlated, most probabiy, with the nutritonal plane, as discussed earlier.
Season influence on open period disappeared when data were aiso classified according to the combination of kidding type, at the beginning of the open period, and at the foilowing kidding.
The longest open period was found for does rai-sing twins and aborting at the following gestations. The shortest open period was observed iii does raising singles and then singies or twins at the foi-lowing kidding. The trend observed was an ex- Pesq. agropec. bras., Brasilia. 23(12):1449-1458, dez. 1988.
TRADITIONAL SYSTEM OF GOAT MANAGEMENT 1457 pected one, even though the analysis was carried
out with smail number of observations. It appears that does kidding twins had a higher number of abortions in the following kidding. This could be related to nutritional plane as well. Animais which aborted in the foliowing kidding were already having some reproductive problerns to conceive, wich is reflected by relativeiy longer open periods. Proper attention to those animais might irnprove their output, mainiy when one considers that about 23% of the conceptions resuited in abor-dons, jeopardizing the profitabiity of the enter-prise.
The high rnorta!ity rate of kids (76.4%) was a striking observation which rnagnifies the ineffi-ciency of the traditional management systern as it was designed for this experiment. Whether the sarne mortahty rates occur at producers' farms remains to be deterrnined. In general, during the dry season, the producers supplement oniy those goats which are iii the worst condition (Primov 1984). This strategy rnight reduce mortaiity rates at farrn levei. In the present study these husban. dry practices were.non existing and may partially explain our observation.
Mortality rates were highest during the dry season period (Fig. 3). Lower birth weights (Lima et ai. 1983), competition for the scarse fodder with aduit animais, and the fact that there are overlapping nursing ages, are some of the causes which rnight have contributed to the high rnortali ty rate.
The highest mortality occurred between seven and 112 days of age (Tabie 5). Mortality rate was not infiuenced by sex or type of birth (Tabie 6). However, season, sex and their interaction had significant effects on the age at death of the 51W kids (Tabie 7). Kids born during the rainy season and males, survived longer than the ones bom during the dry season and femaie kids, respective-ly. Aiso male kids born during the rainy season lived 100 days longer than any of the other clas-ses anaiysed. Figueiredo & Pant (1982) aiso re-ported a significant sex difference, aithough in their case females survived longer. However, since birth weight is positively correiated with survivai period (Figueiredo & Pant 1982) and rnaies showed su-perior birth weights than females under the tra-
ditional system of managemnent (Lima eta1. 1983), it is reasonable to expect that in this study males shouid survive the iongest on the average. High mortahty rates during the preweaning period of about 31% have been reported for Red Sokoto (Adu et ai. 1979), 24% for Ceyion Native (Ra. natunga 1971) and 12.5% for the West African Dwarf (Osuagwuh & Alpokodje 1981) breeds.
In another study at a research station located in Petrolina, PE, Brazii, where comparisons among four goat production systems were carried out, the traditional systern of management showed 32% of kid mortality. The other three improved pro-duction systems showed somewhat iower morta-iity leveis, although a ciear advantage over the tra-ditional system couid not be determined with regard to kid mortality (Padilha et ai. 1980).
CONCLUSIONS
1, An inefficient reproductive performance of native SRD goats mamntained under the traditional systern of rnanagernent as designed under experi-mental farm conditions, as well as a high rate of kid mortality, was found.
2. Some improvement of actual leveis of pro-ductivity couid be achieved by carrying out one breeding season per year, which could be set at any time between September and Decernber de-pending upon the location of the farrn and carefuil historicai survey of precipitation at the particular site. This wouid ensure that the breeding season could start within 100 to 120 days before the beginning of the rainy season.
3. This strategy wiili ailow does to have enough fodder during the iast third of the pregnancy, thus producing heavier kids which couid grow faster and iess prone to death. This approach wili in-crease productivity by reducing kid mortaiity.
4. Other approaches sucli as crossbreeding of native animais with exotic ones to improve proli-flcacy, or the use of supplements at critical periods during the year, remain to be assessed.
1458 E. DE A.M. LIMA eta1.
REFERENCES
ADU, l.F.; BUVANENDRAM, V. LAKPINI, C.A.M. fle reproductive performance of Red Sokoto goats lii Nigeria. J. Agric. Sei., 93(3):563-6, 1979.
BARR, A.L.; GOODNINGHT, J.H.; SALL, J.P.; BLAIR, W.H.; CIIILKQ, D.M. A user's guide to SAS 76. Raleigh, North Caroine, U.S.A., SAS Institute, 1979. 493p.
FIGUEIREDO, E.A.P. & PANT, K.P. Evaivation oígoat breeds in lhe tropical Northeast Brazil. II. An ana-lysis of age at death of kids. Pesq. agropec. bras., 17(5):803-8, 1982.
FORBES, J.M. Factors affecting lhe gestation Iength in sheep. J. Agric. Sei., 68:191-4,1967.
GLIMP, H.A. Effect of breed and mating season on repro-ductive performance of sheep. 1. Anim. Sei., 32(6): 1176-82, 1971.
GONZALEZ-STAGNARO, C. Comportamiento reproduc-tivo caprino en zonas áridas de Venezuela. In: SYMPOSIUM SOBRE LA CABRA EN LOS PAISES MEDITERRANEOS, Málaga, Murcia, Espafia 1977. Proceedings. Madrid, Espafia, Comitá Nacional Es-panhol de Zootecnia, 1977. p.317-21.
GUPTA, 5.; SOM, T.L.; BHATrACHARYA, S. Studies on gestalion length in Black Bengal goats. Indian Vet. J., 41:668-72, 1964.
JARDIM, W.R.; SILVEIRA FILHO, 5.; PEIXOTO, £M.; SPALLINI, A. Observações sobre alguns aspectos da eficiência reprodutiva de caprinos Anglo-nubianos. R. Agric., Piracicaba,40(3):131-47, 1965.
KIRKPATRICIC, R.L. & AICINDELE, Z.T. Reproduction in West African Dwarf goats. J. Anim Sei., 39W: 163, 1974.
LIMA, RAM.; FIGUEIREDO, E.A.P.; SIMPLÍCIO, A.A.; PONCE DE LEÕN, F.A. Traditional system ofgoat management. 1. Pre-weaning growth performance of lhe SRD (non-descript) goats. Pesq. agropec. bras., 18(5):557-63, 1983.
LIMA, F.A.M.; FIGUEIREDO, E.A.P.; SIMPLÍCIO, A.A.; PONCE DE LEÕN, F.A. Traditional system ofgoat management. II. Post-weaning growth performance of lhe SRD (non-descript) goats. Pesq. agropec. bras., 20(3):3 85-93,1985.
MISHRA,R.K.;NIVSARKAR,A.E.;ARORA,L, Anote on lhe analysis of gestation length ia Sirohi goats. Indian J. Anim. Sei., 49(11):967-8, 1979.
OSUAGWUH, A.I.A. & ALPOKODJE, J.U. West African Dwarf (Foulta djallon) goat. I. Causes of early mortality. Int. Goat Sheep Res., 1(14):303-9, 1981. PADILHA, - T.N.; ALBUQUERQUE. S.G. de;
GUIMA-RÃES FILHO, C.; SOARES, J.c.G.; FREIRE, L.C.; SALVIANO, L.M.C.; OLIVEIRA, M.C. de. Compara-ção entre sistemas de produçio para caprinos. Petro-lina, EMBRAPA-CPATSA, 1980. 8p. (EMBRAPA-CPATSA. Pesquisa em Andamento, 8)
PRIMOV, G. Goat production within the farming system of smallliolders of northern Bahia, Brazil. Columbia, 1984, 160p. (SR-CRSP. Technical Report, 35) RANATUNGA, P. lCd mortality on two major
govern-ment farnis in Ceylon. Trop. Anim. Health. Prod., 3:52, 61, 1971.
SACHDEVA, K.K.; SENGAR, O.P.S.; SLNGH, S.N.; LINDAHL, I.L. Studies on goats. I. Effect ofplan of nutrition on the reproductive perfõrmance of does.!. Agric. Sei., 80(3):375-9, 1973.
SHELTON, M. lntluence of the presence of a male goat on lhe initiation of estrous cycling and ovislation af Angora does.!. Anim. Sei., 19(2):368-75, 1960. STEEL, R.G.D. & TORRIE, J.H. PrincipIes and
procedu-res of statistics. New York, McGraw-Hill, 1960. 48lp.
